This 2-year project evaluates the feasibility and efficacy of constrain induced (CI) movement therapy to enhance recovery of motor function in the affected upper limp of patients with ischemic stroke. Motor weakness of the limbs contralateral to the cerebral hemisphere involved in unilateral stroke is a frequent deficit which is a major cause of disability. Animal models and clinical observations have indicated that individuals learn to disuse their paretic limbs and rely more on their intact limbs. Studies in animals and previous clinical trials have shown that constraining the intact limb while training the affected limb improves motor function. However, the feasibility and efficacy of initiating CI Movement therapy during the subacute phase of recovery from ischemic stroke have not been reported. This project is a randomized parallel group design to evaluate the feasibility and efficacy in instituting 2 weeks of daily CI movement therapy combined with training of the impaired upper limb beginning on the eighth day after ischemic stroke. Twenty-eight right handed patients admitted to the Memorial-Herman Hospital Stroke unit who satisfy the selection and exclusion criteria and control condition with the distal motor score of the NIH Stroke Scale and age as stratification variables. Daily training of the impaired limb will also be carried out in both treatment arms during the 2 week period. Two physical therapists will administer all treatment with 1 therapist treating half of the patients in each condition. Outcome measures of upper limb motor function, which include Actual Amount of Use, Motor Activity Log, Grooved Pegboard, Wolf Motor Function Tests, and the motor subscale of the Functional Independence Measure, will be performed by a third physical therapist who is blinded to the patients' treatment condition. In view of animal model studies indicating that lesion volume can enlarge when the intact limb is restrained during the first week after injury, this project includes magnetic resonance imaging (MRI) to measure change in lesion volume from pre-treatment to 3 months post-injury. Reorganization of cortical motor function is studied by functional MRI at 3 months after stroke. The Functional Neuroimaging Core at UCLA will assist with designing the acquisition and analysis of fMRI data. Data management and analysis will be performed in the Cognitive Neuroscience Laboratory with assistance from the Experimental Design and Statistics Core of the University of Washington.